It is useful for diagnosis of cardiac lesion to quantitatively evaluate cardiac output of blood, heart chamber volume, and thickness of cardiac muscle. For this reason, evaluation of these cardiac functions is carried out by medical diagnostic imaging apparatus, such as ultrasonic diagnostic imaging apparatus, magnetic resonance diagnostic imaging apparatus, and X-ray CT diagnostic imaging apparatus.
For example, in diagnosis of cardiac functions using an ultrasonic diagnostic apparatus, movements of valves and cardiac muscle walls, wall thickness, changes of heart chamber volume are measured by setting a tracking point set on a wall of a cardiac muscle and finding out to what location the tracking point has moved in association with pulsation of the heart. This determination of displacement of the tracking point in association with heart pulsation is referred to as tracking of a tracking point, or tracking. There is a method of tracking a tracking point by extracting or setting a small region with an arbitrary shape that contains a tracking point and an adjacent region thereof on each of images and performing a correlation process for this small region between the images. This method is referred to as a template matching technique, and the small region with an arbitrary shape that contains the tracking point and an adjacent region thereof is referred to as a template.
A heart is non-rigid object that undergoes complicated motions and changes in shape complicatedly. Therefore, a tracking point is tracked by a conventional template matching technique, tracking operation error becomes very large. For this reason, the conventional template matching technique has been able to track only one cardiac cycle through three cardiac cycles.
In addition, tracking results of a tracking point by a template matching technique (position coordinates of the tracking point) contains errors originating from sampling, quantization, noise, image quality variations, and examiner's probe manipulation, and these errors accumulate in the course of processing a large number of frames. As a consequence, these errors are added to the tracking point, and a shift from the original position occurs, making it difficult to track the tracking point accurately.
In view of the error problem in each template matching computation, Patent Document 1 discloses a method for improving the accuracy of the tracking operation for a tracking point on a heart, which is a non-rigid object, by varying the size of the template in order to avoid the presence of mixed data of images having the different moving directions of the organ in the same template.    Patent Document 1: JP-A-2004-313291